CN111209119A - Load balancing method for face snapshot rifle bolt - Google Patents
Load balancing method for face snapshot rifle bolt Download PDFInfo
- Publication number
- CN111209119A CN111209119A CN202010071774.9A CN202010071774A CN111209119A CN 111209119 A CN111209119 A CN 111209119A CN 202010071774 A CN202010071774 A CN 202010071774A CN 111209119 A CN111209119 A CN 111209119A
- Authority
- CN
- China
- Prior art keywords
- image
- code rate
- face
- module
- cpu
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 16
- 238000001514 detection method Methods 0.000 claims abstract description 8
- 238000010191 image analysis Methods 0.000 claims abstract description 4
- 230000003247 decreasing effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract 1
- 230000005540 biological transmission Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
- G06F9/505—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals considering the load
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K17/00—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations
- G06K17/0022—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations arrangements or provisions for transferring data to distant stations, e.g. from a sensing device
- G06K17/0025—Methods or arrangements for effecting co-operative working between equipments covered by two or more of main groups G06K1/00 - G06K15/00, e.g. automatic card files incorporating conveying and reading operations arrangements or provisions for transferring data to distant stations, e.g. from a sensing device the arrangement consisting of a wireless interrogation device in combination with a device for optically marking the record carrier
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Computer Networks & Wireless Communication (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of face recognition, in particular to a load balancing method of a face snapshot rifle bolt, which comprises the steps of collecting two paths of original images, and transmitting the original images to an image coding and decoding module by a collection module of the snapshot rifle bolt; the image coding and decoding module outputs one path of image to the CPU coding module for coding, and outputs the other path of image to the GPU module for face detection; after receiving the image, the CPU image application module stores the image to a local memory card and uploads the image to a remote server through a network; the CPU image analysis module analyzes the local storage image, counts the number N of face snapshots within a set time, transmits a code rate value X to be adjusted, and acquires and configures an initial code rate value as the code rate value X by the CPU coding module; the coding code rate is dynamically adjusted in real time, the code rate is reduced when the snapshot amount is large, CPU resources are made available for local storage of pictures, and meanwhile, the effects of uploading the pictures to a remote server are achieved; otherwise, the code rate can be recovered and even improved in a period with small snapshot amount, and a load balancing strategy is achieved.
Description
Technical Field
The invention relates to the technical field of face recognition, in particular to a face snapshot rifle bolt load balancing method.
Background
In recent years, with the popularization of artificial intelligence chips, the face snapshot rifle bolt gradually becomes an indispensable product in the security field, and in view of the limitation of embedded device resources (such as memory, GPU and CPU processing capability), how to fully utilize the limited resources to exert the maximum performance of the face snapshot rifle bolt becomes a key technology. The face snapshot rifle bolt is characterized in that face snapshot is carried out, and equipment resources of the face snapshot rifle bolt need to preferentially ensure local storage of a snapshot picture and remote server uploading of the snapshot picture.
In conventional bolt face applications, the encoding rate is usually limited to static parameters, and the installation and debugging personnel can determine an initial value by experience in combination with factors such as application scenarios and network bandwidth. If the fixed and unchangeable code rate is set to be smaller, the video stream is not clear and the equipment resources are wasted; if the setting is larger, the frame loss of the image and the overload work of the equipment can be caused, on the basis, how to dynamically adjust the coding code rate in real time by counting the current face snapshot number can be realized, the code rate can be reduced in a time period with large snapshot amount, the CPU resource can be saved for the local storage of the image, and the network bandwidth can be saved for the uploading of the image to a remote server; otherwise, the code rate can be recovered and even improved in a period with small snapshot amount, and a load balancing strategy is achieved.
Disclosure of Invention
The invention aims to solve the technical problem that the coding code rate is dynamically adjusted in real time by counting the current face snapshot number, so that the current face snapshot number is in a load balancing state.
The technical scheme adopted by the invention is as follows: the face snapshot rifle bolt load balancing method comprises the following steps:
step S1, collecting two paths of original images, and transmitting the original images to an image coding and decoding module by a collection module of the snapshot rifle bolt;
step S2, after receiving two paths of original images, the image coding and decoding module outputs one path to the CPU coding module for coding, and outputs the other path to the GPU module for face detection;
step S3, the CPU image application module receives the image coded by the CPU coding module and the face detection of the GPU module, stores the image to a local memory card, and uploads the image to a remote server through a network;
step S4, the CPU image analysis module analyzes the local storage image, counts the face snapshot number N in the set time, and supposes that the initial code rate value is M, the code rate value is needed to be adjusted to be X:
X=M * 1/([N/500]+1)
and step S5, transmitting the code rate value X to be adjusted, and the CPU encoding module acquires and configures the initial code rate value as the adjusted code rate value X.
Further, in step S2, if no human face is detected, step S2 is continuously executed until a human face is detected.
Further, in step S4, when the number of face snapshots is greater than the set value, the current code rate value is decreased according to the initial code rate.
Further, in step S4, when the number of face snapshots is less than the set value, the current code rate value is increased according to the initial code rate.
Further, in step S4, the statistical setting time is set to 10 minutes.
The invention has the beneficial effects that:
1) in the face snapshot rifle bolt, a CPU is responsible for work such as coding and decoding, network transmission and the like, an artificial intelligent chip GPU is responsible for face detection, when the flow of people is large, the power of the GPU is increased, pictures of the snapshot face are increased, the workload of the CPU is correspondingly increased at the moment, a local picture storage and uploading platform occupies a large amount of CPU resources, at the moment, the coding code rate is dynamically reduced, the snapshot efficiency can be effectively improved, and valuable CPU resources and network resources are used for local storage and remote transmission of the pictures.
2) The method can realize dynamic and reasonable code stream change, solve the contradiction between definition and code rate under simple scenes and complex scenes, reduce the occupation of network bandwidth, balance the load of a CPU and ensure the quality of face snapshot images.
3) The dynamic code rate adjusting method overcomes the defect that the former installer fixedly adjusts the code rate by experience, can be self-adapted to various scenes such as residential building corridors, roads, subways, squares and the like, and has wide application prospect.
Drawings
Fig. 1 is a schematic flow chart of a load balancing method for a face snapshot bolt of the present invention.
Detailed Description
The invention is further described below with reference to the accompanying drawings.
The CPU needs to be highly versatile to handle various data types, while requiring logical decisions that introduce significant branch jumps and interrupts. These all make the internal structure of the CPU exceptionally complex. GPUs are faced with a clean computing environment of highly uniform type, independent of large-scale data, and need not be interrupted. Unlike CPUs which are good at logic control and general-purpose type data operation, GPUs are good at large-scale concurrent computation, which is also required for password cracking and the like. This scheme selects a GPU for a large amount of image processing. And selecting a CPU for coding and calculating.
The scheme of the application mainly relates to a load balancing method for a face snapshot bolt, which comprises the following steps:
s1, collecting two paths of original images, and transmitting the original images to an image coding and decoding module by a collection module of the snapshot rifle bolt;
s2, after receiving two paths of original images, the image coding and decoding module outputs one path of original images to a CPU coding module for coding, and outputs the other path of original images to a GPU module for face detection;
s3, the CPU image application module receives the image coded by the CPU coding module and the face detection of the GPU module, stores the image to a local memory card, and uploads the image to a remote server through a network;
s4, analyzing the local storage image by the CPU image analysis module, counting the number N of face snapshots within a set time, and if the initial code rate value is M, adjusting the code rate value to X:
X=M * 1/([N/500]+1)
and S5, transmitting the code rate value X to be adjusted, and the CPU encoding module acquires and configures the initial code rate value as the adjusted code rate value X.
In S2, if no face is detected, step 2 is continued until a face is detected.
In step S4, when the number of face snapshots is greater than the set value, the current code rate value is decreased according to the initial code rate.
In step S4, when the number of face snapshots is less than the set value, the current code rate value is increased according to the initial code rate.
In step S4, the statistical setting time is set to 10 minutes.
The embodiment adopts a camera to capture face image files of the faces of people to be detected or obtain the existing photos to form the face image files, and generates face print codes for storing the face image files through a CPU coding module. And the other path is subjected to face recognition or photo input, and then the two paths of processed images are stored in a local server and a remote server.
Another scheme is to create a face file and compare the current face print code with the file stock. The face print code of the current face image is searched and compared with the face print code in the file stock. The above coding mode works mainly according to the essential characteristics of human faces. The code has a strong reliability so that it can accurately identify a person from a group of people. The human face recognition process can be automatically, continuously and in real time completed by using common image processing equipment. The method can realize dynamic and reasonable code stream change, solve the contradiction between definition and code rate under simple scenes and complex scenes, reduce the occupation of network bandwidth, balance the load of a CPU and ensure the quality of face snapshot images.
Claims (5)
1. A load balancing method for a face snapshot rifle bolt is characterized by comprising the following steps:
step S1, collecting two paths of original images, and transmitting the original images to an image coding and decoding module by a collection module of the snapshot rifle bolt;
step S2, after receiving two paths of original images, the image coding and decoding module outputs one path to the CPU coding module for coding, and outputs the other path to the GPU module for face detection;
step S3, the CPU image application module receives the image coded by the CPU coding module and the face detection of the GPU module, stores the image to a local memory card, and uploads the image to a remote server through a network;
step S4, the CPU image analysis module analyzes the local storage image, counts the face snapshot number N in the set time, and supposes that the initial code rate value is M, the code rate value is needed to be adjusted to be X:
X=M * 1/([N/500]+1)
and step S5, transmitting the code rate value X to be adjusted, and the CPU encoding module acquires and configures the initial code rate value as the adjusted code rate value X.
2. The load balancing method for the face snapshot bolt according to claim 1, characterized in that: in step S2, if no face is detected, step 2 is continued until a face is detected.
3. The load balancing method for the face snapshot bolt according to claim 1, characterized in that: in step S4, when the number of face snapshots is greater than the set value, the current code rate value is decreased according to the initial code rate.
4. The load balancing method for the face snapshot bolt according to claim 1, characterized in that: in step S4, when the number of face snapshots is less than the set value, the current code rate value is increased according to the initial code rate.
5. The load balancing method for the face snapshot bolt according to claim 1, characterized in that: in step S4, the statistical setting time is set to 10 minutes.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010071774.9A CN111209119A (en) | 2020-01-21 | 2020-01-21 | Load balancing method for face snapshot rifle bolt |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010071774.9A CN111209119A (en) | 2020-01-21 | 2020-01-21 | Load balancing method for face snapshot rifle bolt |
Publications (1)
Publication Number | Publication Date |
---|---|
CN111209119A true CN111209119A (en) | 2020-05-29 |
Family
ID=70788459
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010071774.9A Pending CN111209119A (en) | 2020-01-21 | 2020-01-21 | Load balancing method for face snapshot rifle bolt |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111209119A (en) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111741305A (en) * | 2020-08-27 | 2020-10-02 | 科大讯飞(苏州)科技有限公司 | Video coding method and device, electronic equipment and readable storage medium |
CN112565762A (en) * | 2020-12-04 | 2021-03-26 | 上海航天计算机技术研究所 | Method and device for equalizing coding of multichannel video images suitable for carrier rocket |
CN114845119A (en) * | 2022-07-04 | 2022-08-02 | 光谷技术有限公司 | Thing allies oneself with gateway and verifies and compression system |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN201733373U (en) * | 2010-05-07 | 2011-02-02 | 北京英泰智软件技术发展有限公司 | High definition face capturing distinguishing camera |
US20110211036A1 (en) * | 2010-02-26 | 2011-09-01 | Bao Tran | High definition personal computer (pc) cam |
US20150022666A1 (en) * | 2013-07-22 | 2015-01-22 | Intellivision Technologies Corp. | System and method for scalable video cloud services |
CN108376246A (en) * | 2018-02-05 | 2018-08-07 | 南京蓝泰交通设施有限责任公司 | A kind of identification of plurality of human faces and tracking system and method |
CN109040587A (en) * | 2018-08-01 | 2018-12-18 | 北京旷视科技有限公司 | It captures processing method, device, capture mechanism, equipment and storage medium |
CN109241111A (en) * | 2018-08-27 | 2019-01-18 | 武汉虹信技术服务有限责任公司 | A kind of distributed face identification system and method for database based on memory |
CN109818789A (en) * | 2019-01-22 | 2019-05-28 | 苏州科达科技股份有限公司 | Picture transmission method and system, storage medium, electronic equipment |
CN110493534A (en) * | 2019-05-31 | 2019-11-22 | 杭州海康威视数字技术股份有限公司 | The method and apparatus of Image Acquisition and processing |
CN110648225A (en) * | 2018-06-26 | 2020-01-03 | 玛链(上海)网络技术有限公司 | Method for applying digital code chain |
-
2020
- 2020-01-21 CN CN202010071774.9A patent/CN111209119A/en active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20110211036A1 (en) * | 2010-02-26 | 2011-09-01 | Bao Tran | High definition personal computer (pc) cam |
CN201733373U (en) * | 2010-05-07 | 2011-02-02 | 北京英泰智软件技术发展有限公司 | High definition face capturing distinguishing camera |
US20150022666A1 (en) * | 2013-07-22 | 2015-01-22 | Intellivision Technologies Corp. | System and method for scalable video cloud services |
CN108376246A (en) * | 2018-02-05 | 2018-08-07 | 南京蓝泰交通设施有限责任公司 | A kind of identification of plurality of human faces and tracking system and method |
CN110648225A (en) * | 2018-06-26 | 2020-01-03 | 玛链(上海)网络技术有限公司 | Method for applying digital code chain |
CN109040587A (en) * | 2018-08-01 | 2018-12-18 | 北京旷视科技有限公司 | It captures processing method, device, capture mechanism, equipment and storage medium |
CN109241111A (en) * | 2018-08-27 | 2019-01-18 | 武汉虹信技术服务有限责任公司 | A kind of distributed face identification system and method for database based on memory |
CN109818789A (en) * | 2019-01-22 | 2019-05-28 | 苏州科达科技股份有限公司 | Picture transmission method and system, storage medium, electronic equipment |
CN110493534A (en) * | 2019-05-31 | 2019-11-22 | 杭州海康威视数字技术股份有限公司 | The method and apparatus of Image Acquisition and processing |
Non-Patent Citations (1)
Title |
---|
李昀 等: "人脸识别的技术分析及项目实践", 《中国电信业》 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111741305A (en) * | 2020-08-27 | 2020-10-02 | 科大讯飞(苏州)科技有限公司 | Video coding method and device, electronic equipment and readable storage medium |
CN111741305B (en) * | 2020-08-27 | 2020-11-24 | 科大讯飞(苏州)科技有限公司 | Video coding method and device, electronic equipment and readable storage medium |
CN112565762A (en) * | 2020-12-04 | 2021-03-26 | 上海航天计算机技术研究所 | Method and device for equalizing coding of multichannel video images suitable for carrier rocket |
CN112565762B (en) * | 2020-12-04 | 2022-08-12 | 上海航天计算机技术研究所 | Method and device for equalizing coding of multichannel video images suitable for carrier rocket |
CN114845119A (en) * | 2022-07-04 | 2022-08-02 | 光谷技术有限公司 | Thing allies oneself with gateway and verifies and compression system |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111209119A (en) | Load balancing method for face snapshot rifle bolt | |
CN109801230B (en) | Image restoration method based on encoder structure | |
WO2012149296A2 (en) | Providing content aware video adaptation | |
CN111479048A (en) | Intelligent video image processing equipment based on edge calculation | |
CN115914649A (en) | Data transmission method and system for medical video | |
CN112565777B (en) | Deep learning model-based video data transmission method, system, medium and device | |
CN104041016A (en) | Camera device, server device, image monitoring system, control method of image monitoring system, and control program of image monitoring system | |
CN104320622A (en) | Embedded video enhancement system for open source server software | |
CN103700121A (en) | Method and device for compressing composite image | |
CN110557633A (en) | Compression transmission method, system and computer readable storage medium for image data | |
Raya et al. | Analysis realization of Viola-Jones method for face detection on CCTV camera based on embedded system | |
US20230023369A1 (en) | Video processing method, video processing apparatus, smart device, and storage medium | |
Zhu et al. | A semantic-aware transmission with adaptive control scheme for volumetric video service | |
CN107566785B (en) | Big data-oriented video monitoring system and method | |
Hou et al. | Real-time surveillance video salient object detection using collaborative cloud-edge deep reinforcement learning | |
Murad et al. | Dao: Dynamic adaptive offloading for video analytics | |
Zhang et al. | Mfvp: Mobile-friendly viewport prediction for live 360-degree video streaming | |
CN117221609B (en) | Centralized monitoring check-in system for expressway toll service | |
Mei et al. | Learn a compression for objection detection-vae with a bridge | |
CN113191945A (en) | High-energy-efficiency image super-resolution system and method for heterogeneous platform | |
CN104301722A (en) | Video stream fuzzy detection method based on frequency domain | |
CN116709037A (en) | 5G camera-based method for transmitting monitoring key data of weather and rain | |
CN115984675A (en) | System and method for realizing multi-channel video decoding and AI intelligent analysis | |
Ikeda et al. | Reduction of communication cost for edge-heavy sensor using divided cnn | |
CN113810695A (en) | Video encoding method, apparatus and computer-readable storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200529 |